登录 EN

添加临时用户

房价预期、居民消费与疫情冲击——基于文本数据

Housing Price Expectations, Household Consumption, and the Impact of COVID-19: A Text-Based Analysis

作者:赵城敏
  • 学号
    2020******
  • 学位
    硕士
  • 电子邮箱
    nat******com
  • 答辩日期
    2025.05.12
  • 导师
    刘涛雄
  • 学科名
    理论经济学
  • 页码
    87
  • 保密级别
    公开
  • 培养单位
    070 社科学院
  • 中文关键词
    房价预期;居民消费;预期理论;诱导预期;疫情冲击
  • 英文关键词
    House Price Expectations; Household Consumption; Expectation Theory; Heuristic Expectation; Pandemic Shock

摘要

从中国宏观经济特征来看,房价高速增长已经成为影响居民消费支出行为的一个不可忽视的因素。中国城镇家庭中有74.2%的实物资产集中于房地产,因此,家庭对房价的预期变化会影响其对财富积累路径的判断,进而对消费与储蓄行为产生影响。扭曲居民经济行为机制的核心因素并非是房价本身,而是对房价的预期。因此,更加基于现实的对家庭房价预期的研究,对于中国经济转型问题以及构建“双循环”新发展格局具有重要意义。但是,目前在学界从预期理论的角度分析房价预期和家庭消费的文献较少,而且已有文献没有克服适应性预期和理性预期的框架,对有限理性的讨论却有所不足的水平。对此,本文将传统预期理论和行为经济学的有限理性启发法相结合,把家庭房价预期分为理性预期、适应性预期和以社交媒体文本数据为基础的诱导预期(Heuristic Expectation),建立预期方程和消费函数,分析家庭对房价的不同预期对居民消费产生的影响。为此,本文搜集新闻、论坛、微信、微博的主贴/回帖文本数据,而对此进行基于机器学习(Machine Learning)的情感分析,获取关于房价的高质量文本数据。进一步,本文还研究在2020年发生的史无前例的经济危机——新冠病毒肺炎疫情所带来的冲击,将分析疫情发生了之后,不同预期对家庭消费产生的影响如何变化。最后,基于此,提出政策建议。

From a macroeconomic perspective, surging housing prices have become a key factor shaping household consumption in China. With 74.2% of urban household assets tied up in real estate, shifts in housing price expectations significantly influence perceptions of wealth, thereby affecting consumption and savings behavior. Crucially, it is these expectations—not prices themselves—that distort household decision-making. Therefore, a more realistic and nuanced understanding of household housing price expectations is essential for addressing China’s economic transition and advancing the “Dual Circulation” development strategy. However, existing academic research offers limited analysis of housing price expectations and household consumption from the perspective of expectation theory. Moreover, prior studies often remain confined within the frameworks of adaptive or rational expectations, with insufficient exploration of bounded rationality. To address this gap, this paper integrates traditional expectation theory with insights from bounded rationality and behavioral economics (heuristic theory), categorizing household housing price expectations into three types: rational expectations, adaptive expectations, and externally-guided expectations based on social media text data. Based on this framework, the study constructs expectation equations and consumption functions to analyze the impact of different types of expectations on household consumption. To achieve this, the study collects a wide range of textual data—including news articles, Luntan, WeChat, and Weibo posts/comments—and applies machine learning-based sentiment analysis to extract high-quality textual indicators related to housing prices. Furthermore, the paper examines the impact of the unprecedented economic crisis brought on by the COVID-19 pandemic in 2020, analyzing how the effects of different expectation types on household consumption evolved in the aftermath of the outbreak. Based on the findings, the paper offers corresponding policy recommendations.